Companies Are the Real School
I’ve read the books. They help. But the most useful education I’ve gotten didn’t come from a book, a podcast, or a thread. It came from a customer who left, a payroll that had to clear, a hire who didn’t work out, and a month where the math didn’t. A company turns out to be a school — one where reality grades your thinking every day, in public, with money.
This is the companion to the charter. That essay explains why I publish. This one explains why I trust what I publish: it’s graded work, not commentary.
Theory gives you language. Operating gives you judgment.
Books are good at one thing in particular: they give you language. Naming a thing — cost disease, the bullwhip effect, the principal-agent problem — is genuinely useful, because you can’t manage what you can’t name. I’m not anti-theory. Half of what’s on the Library shelf earned its place by handing me a word for something I’d felt but couldn’t describe.
But language is not judgment. Judgment is knowing which of the named things is actually happening to you right now, how bad it is, and what it will cost if you’re wrong. You don’t get that from the definition. You get it from being on the hook when it’s live. A commentator can describe ten failure modes fluently and has felt none of them. The operator has felt three of them this quarter and is trying not to feel the fourth. Those are different kinds of knowing, and only one of them holds up when there’s money on it.
Reality is stricter than theory
The thing about a real company is that it exposes weak assumptions fast, and it doesn’t grade on a curve. A plan that sounds airtight in a document meets a customer who never read the document. A pricing model that’s elegant on a whiteboard meets a buyer who shrugs. A process that works at ten clients quietly breaks at fifty.
Theory lets you round off the inconvenient variable. Operations sends you the invoice for it. Every assumption I held loosely got tested whether I wanted it tested or not, and the wrong ones got expensive. That strictness is the feature, not the bug — it’s exactly what makes the lesson stick. You remember the assumption that cost you a month far longer than the one a book warned you about.
The teachers
Once you’re inside it, you notice a company isn’t one teacher. It’s a faculty, and each member teaches a different subject.
Customers teach what the market actually values — which is almost never what you decided it should value. You can argue with a strategy deck. You cannot argue with a customer who doesn’t renew. They vote with attention, money, and repeat business, and the vote is final.
Cash flow teaches discipline faster than any advice can. Payroll leaves on the 1st; the invoice arrives on the 45th; and the gap between those two dates has taught more founders more real economics than any course — a lesson sharp enough that it gets its own entry. Money in motion is a stricter tutor than money in theory.
Hiring teaches that people are not abstractions. An “org chart” is a diagram; a team is specific humans with their own mornings, limits, and judgment. You learn — sometimes the hard, honest way — that the people decisions are the ones that compound the longest, in both directions.
Operations teach that execution is not beneath strategy — it’s where strategy becomes real or doesn’t. The deck is a hypothesis. The resolved ticket, the completed payroll run, the closed month: that’s the experiment. Strategy that never survives execution wasn’t strategy. It was a wish with good formatting.
Mistakes are the faculty’s most effective members. They charge tuition, and the bill clears. The cost isn’t the problem — that’s just how this school takes payment. The only real waste is paying the same tuition twice for the same lesson, which is most of what writing things down is for.
Everyone else — vendors, competitors, regulators, the market itself — teaches systems and power. A vendor’s incentives are not yours. A competitor moves whether or not your plan accounted for it. A regulation can reshape the board mid-game. You learn to read situations by their incentives rather than their press releases, because the people across the table are running their own math, not yours. That habit — incentives first, narrative second — is the closest thing to a master key this school hands out.
Four classrooms
I’m not enrolled in one school. The companies I’m building each teach a different curriculum, and the overlap is where it gets interesting. (The current roster is on the Companies page.)
Defrilex — multilingual customer support and interpretation — teaches service: client trust, multilingual operations, and the unglamorous reality of workflow, where quality is decided in handoffs nobody sees.
Prolify — finance software for founders and multi-company operators — teaches money: cash flow, multi-entity visibility, and the operating discipline that only shows up when the numbers actually have to reconcile.
AI Thinking Lab teaches applied intelligence: what these systems really do once you deploy them, where automation earns its place, and where judgment still has to sit in the loop.
Vectis — global payroll, EOR, and workforce infrastructure — teaches employment at scale: compliance, cross-border complexity, and the truth that paying people correctly, everywhere, on time is a promise before it is a feature.
Four classrooms, one student. The same forces show up in all of them wearing different clothes — and learning to see the pattern across them, rather than treating each as its own story, is most of what I’m actually trying to do.
Why the Lessons section exists
This is why there’s a whole section of the site called Lessons, and why it isn’t a feed of generic advice. Generic advice is what you write when you haven’t paid for the lesson. The entries there are the things building actually charged me for, recorded with the specifics intact, so the tuition buys something permanent instead of evaporating.
The test for what belongs is simple: if a piece of advice could have been written by someone who never ran the thing, it doesn’t go in. The point of the section is to keep the receipts.
The goal is to become harder to fool
I want to be clear about the goal, because it’s easy to mistake. The goal of all this is not to look smart in public. Looking smart is cheap, and getting cheaper. The goal is to become harder to fool — about markets, about people, about my own reasoning — and a company is the most honest instrument I’ve found for that, because it returns a verdict whether or not I like it.
Publishing is what makes the verdicts compound. A lesson learned and not written down gets relearned at full price. Written down, dated, and put in public, it stops being a private scar and becomes a position I can build on — and be corrected on. That’s the whole machine: build, get graded, write it down, do better, repeat, in public.
The classroom never closes. That’s the part I like.